Unlocking the Potential: How Trustworthy AI Reshapes Industries
Risks and Ethical Considerations: The Unchartered Terrain
Many risks and ethical considerations emerge as the inevitable shadows of its brilliance in this burgeoning frontier of Generative AI. Some of the risks and limitations associated with Generative AI include:1) Hallucination: A scenario when a false/out-of-context response is generated to a query. This response could be grammatically or sometimes even semantically incorrect. This is due to pre-trained results.2) Token Size Limits: The maximum token size limit per model has increased. It is 16K (GPT 3.5 Turbo) in a single call, which can make it more challenging to process larger documents.3) Adversarial Behaviour: It is critical to minimise risk proactively, which can be from malicious behaviour on the network. This is necessary to maintain operations and customer trust.4) Ethical Use: Organisations must ensure that AI usage aligns with the purpose of the overall exercise, and it is imperative to assess manual intervention in the loop when evaluating AI’s suggestions. Moreover, generative AI systems harbour the potential to fabricate sensitive personal data, such as medical records and financial information. The exploitation of this data could fuel identity theft and other criminal activities. Thus, the onus lies on stakeholders to diligently assess these risks and proactively mitigate them.
Addressing the Risks
India’s Newly Enacted Data Privacy Law
Charting the Path Forward
The imperative for trustworthy AI frameworks intensifies as the prominence of Generative AI continues to ascend. To navigate this evolving landscape successfully, several pivotal factors warrant incorporation:1) International Collaboration: Fostering collaborative global partnerships to develop and refine trustworthy AI frameworks.2) Continual Enhancement: Commitment to perpetual evaluation and refinement of these frameworks in tandem with the evolving Gen AI technology.3) Public Education: Elevating public awareness concerning the potential advantages and risks of Generative AI, enabling informed decision-making.Embracing these considerations, organisations can stride confidently into the future, propelled by the remarkable potential of Generative AI while safeguarding ethical integrity and societal well-being.About authors: Sathish Gopalaiah is the president, consulting at Deloitte South Asia, and Payal Agarwal is partner, consulting at Deloitte India.
